The present invention is related generally to the field of optical component alignment in an optical assembly and, more particularly, to arrangements and an associated method for aligning and fixedly supporting an optical component, as part of an optical assembly, with high precision in relation to an optical path.
The prior art appears to be replete with approaches for supporting optical components in an aligned condition, for example, within an optical package. While it is admitted that a great number of these approaches are at least generally effective, it is submitted in view of the discoveries brought to light herein, that these approaches appear to share a number of heretofore unresolved problems and disadvantages, as will be further described.
One common approach of the prior art, with respect to optical component positioning, relies on supporting an optical component on a “clip”. The latter is typically attached to a support surface which is defined within the optical package. One, more recent implementation of this approach is seen in U.S. Pat. No. 6,184,987, issued to Jang et al (hereinafter, Jang). The clip of Jang supports, for example, an optical fiber. A metal ferrule supports the end of the fiber. The clip is laser welded to a support surface and the ferrule is laser then welded to the clip. Laser hammering is then employed in a way that is intended to compensate for weld shifts. Since the patent contemplates sub-micron positioning, the need for further alignment adjustments is virtually assured. While various techniques are available in attempting to achieve precise positioning, Jang utilizes laser hammering, in which additional welds are made, thereby attempting to induce strains in strategic locations to bend the mounting fixture and thereby move the fiber end back into a desired position. Unfortunately, it is not possible to precisely predict or control hammering induced strains. This method, like any method using laser hammering, therefore, inevitably relies on some level of trial and error. For this reason, the technique introduced by Jang is likely, at best, to be time consuming and is at least potentially unreliable. It is also important to note that laser hammering can result in residual stresses that relax over time, as the unit is subjected to temperature cycling in everyday storage and use. The outcome may be a mounting arrangement that exhibits long-term creep with an attendant performance degradation over time.
Another approach taken by the prior art also relies on the use of a clip to support the optical component. Unlike the approach exemplified by the '987 patent, however, a positioning arrangement is used both to move the optical component into a desired position and to then hold the component in the desired position as the clip is welded in place within the optical assembly. This approach may be referred to in a generic sense as “direct-clamping.” Generally, the term direct-clamping, as used herein, refers to any arrangement wherein a clamp or holding tool manipulates the optical component. In this regard, a clip that is suitable for use in direct-clamping may take on any number of configurations. In one form of direct-clamping, the support clip is at least somewhat spring-like. Unfortunately, however, this implementation is subject to relatively large magnitudes of spring-back upon release of the component. While spring-back is generally problematic using any form of direct-clamping, it is submitted that weld shift still further exacerbates the spring-back problem since the flexible clip absorbs much of the weld shift in the form of elastic deformation. Release of the flexible clip serves to release the absorbed elastic deformation thereby resulting in movement of the optical component. In sum, the offending stresses which produce spring-back are difficult to avoid, and unless the mounting clip is extremely rigid, small stresses may lead to large shifts of up to 3–10 microns. One approach seen in the prior art in an attempt to cope with this is to use calibrated or calculated overextension. That is, moving past the desired position prior to release, and/or, in another approach, by performing laser-hammering after release. Unfortunately, it is submitted that these more traditional approaches at best are touchy, process-sensitive and potentially time consuming.
Another form of direct-clamping, referred to herein as “hard-clamping”, is specifically intended to overcome the problem of such weld shifts inducing a corresponding positional shift in the optical component. At the same time, this method attempts to cope with the spring-back problem, described above, particularly complicated through the use of a flexible clip. Accordingly, the positioning arrangement used in a hard-clamping implementation must be capable of maintaining sufficient force (roughly ten to thirty pounds of force), to rigidly hold the optical component in place while welding is performed. This method raises concerns in requiring a bulky, rigid clamp, as well as a rigid component or clip (typically, the part to be mounted consists of solid machined construction). These structural mandates are imposed for the purpose of supporting the optical component to resist or overcome weld-induced forces which tend, in turn, to produce the subject unpredictable changes in positional relationships. Another concern is introduced wherein the optical component is of insufficient strength to endure this form of direct-clamping. In this instance, the method may be implemented by providing a rigid support structure having a mounting platform with the optical component mounted thereon. The mounting platform is then itself hard-clamped to resist weld shift. Unfortunately, it is submitted that hard-clamping techniques, in any of the described varieties, encounter significant difficulties in attempting to produce reliable positioning (as contemplated herein, to ±0.1 um tolerance). In order to reach the contemplated degree of precision, subsequent post-weld hammering or bending is typically needed. In addition to the foregoing concerns, it should be appreciated that in any technique using some form of clamp or holder to position the optic (or its directly supporting platform) before and during the welding process, the clamping tool must be disengaged at some point. Unclamping unavoidably releases residual forces, thereby causing at least some undesired spring-back such that this problem remains unresolved.
As alluded to above, another common approach in the prior art involves post-weld bending. That is, after initial positioning and welding, a holding clamp is used to bend support members, such as legs, which support the optical component, to move the optical component into the desired position. Often, the support members are designed specifically for quick onset of plastic deformation as bending forces are applied. This approach, however, shares a disadvantage with direct-clamping. Specifically, some level of elastic spring-back will typically follow any attempt at precision bending. While spring-back can be compensated for somewhat predictably by intentionally overshooting the desired position, it is submitted to be extremely difficult to compensate to within 0.1 um tolerances. In this regard, one of the attractions of laser hammering, in contrast to post weld bending and direct-clamping, resides in the fact that the holding tool is disengaged prior to the fine-adjustment steps, so there is no tool removal-induced spring-back.
A more recent approach to the problem of weld shifts is demonstrated by FIG. 4A of Published International Patent Application No. WO 01/18580 by Webjorn et al (hereinafter Webjorn). The subject figure illustrates a support structure having an elongated main body including a pair legs positioned proximate to either end. The optical component is described as being positioned “close” to a front pair of the legs. The main body includes a pair of gripping holes arranged proximate to each of its ends. This structure is used by performing an initial alignment using a gripping tool which engages the pair of gripping holes closest to the optical component. Following this initial alignment, the front pair of legs is welded. A second alignment step compensates for “post-attachment shift” by gripping the main body using the pair of gripping holes at its rear end. The structure is described as allowing a “small positioning correction” to compensate for weld shift produced by attachment of the front legs through moving the rear of the support structure. The rear legs are then welded and the gripping tool is removed. Of course, a weld shift is also produced upon welding of the rear legs, however, based on this configuration, the weld shift at the rear of the structure would be expected to produce a corresponding, but reduced magnitude of shift at the optical component.
Still considering Webjorn, while the described support structure and technique should be generally effective in achieving precision alignment, it is submitted that important disadvantages accompany its use. It is submitted that the very length of the elongated main body along the light path and which is required in order to achieve precision movement, already renders the device far too long for many applications. More importantly, any bending or other such distortion, for example, due to thermal stresses or mounting, in the overall package which houses the Webjorn arrangement will result in reduced optical coupling. It is herein recognized that the length of the main body, even without an increase therein for purposes of adjustment enhancement, is likely to disadvantageously require an overall package outline that is stiffer and bulkier (and therefore larger and more costly) than would traditionally be required. While the Webjorn disclosure describes a shorter, two-legged clip, the precision alignment technique is applicable only to a four-legged structure.
At first blush, it may appear that increasing the main body length of Webjorn and, hence, the separation distance between the front and rear pairs of legs is attended only by advantages in further refining adjustment precision. Any resultant advantage, however, is at cross-purposes with other objectives, inasmuch as miniaturization is a substantial motivation in producing many optical assemblies. That is, an increase in length still further complicates matters with respect to package bending and outline.
Another recent approach is seen in U.S. Pat. No. 5,833,202 issued to Wolfgang. The latter introduces a tripod-like component support structure which is intended to be deformable or bendable for positioning adjustments. With respect to precision alignment, however, Wolfgang is subject to weld shift and spring-back effects as a result of its apparent reliance on direct-clamping, which is described in terms of micro-manipulation of the mounted optical component at column 6, lines 54–55. Moreover, it is submitted that the sole structure described in detail by Wolfgang, a tripod, is not well-suited for linear stacking along an optical path for purposes of producing a compact assembly. In this regard, it is noted that this disadvantage is shared with Webjorn since the latter requires the use of an elongated main body.
The present invention resolves the foregoing disadvantages and problems while providing still further advantages, as will be described.